Discover MakerZone

MATLAB and Simulink resources for Arduino, LEGO, and Raspberry Pi

Learn more

Discover what MATLAB® can do for your career.

Opportunities for recent engineering grads.

Apply Today

New to MATLAB?

Thread Subject:
Please Help Me To Crop This hand signal on this region?

Subject: Please Help Me To Crop This hand signal on this region?

From: john john

Date: 7 Feb, 2012 13:51:09

Message: 1 of 1

Please Help Me To Crop This hand signal on this region? I am doing hand signal recognition, I already track the hand, make it black and white the image, I already mark the main image to be crop.. now what I want is to control region being crop in the image..what i want is to crop the region shown in the image(i put a line)....(see this links below) How do I do that? refer the Code Below...
input image =
output image =

Refer the code below I am using..
    x = imread('C:\Users\Jonas Mamaran\Documents\MATLAB\pray\a\akugangsss.jpg');
            rawImage = x;
            I = rawImage;
            cform = makecform('srgb2lab');
            J = applycform(I,cform);
            ab = double(J(:,:,2:3));
            nrows = size(ab,1);
            ncols = size(ab,2);
            ab = reshape(ab,nrows*ncols,2);
            nColors = 2;
            [cluster_idx cluster_center] = kmeans(ab,nColors,'distance','sqEuclidean', ...
            pixel_labels = reshape(cluster_idx,nrows,ncols);
            segmented_images = cell(1,2);
            rgb_label = repmat(pixel_labels,[1 1 3]);
            for k = 1:nColors
            color = I;
            color(rgb_label ~= k) = 0;
            segmented_images{k} = color;
            mean_cluster_value = mean(cluster_center,2);
            [tmp, idx] = sort(mean_cluster_value);
            hand_cluster_num = idx(2);
            K = segmented_images{hand_cluster_num};
            L = im2bw(segmented_images{hand_cluster_num},0.001);
            s = regionprops(rgb_label, I, {'Centroid','BoundingBox'});
            % s(hand_cluster_num).Centroid
            % s(hand_cluster_num).BoundingBox
            centroid_x = uint16(s(hand_cluster_num).Centroid(1));
            centroid_y = uint16(s(hand_cluster_num).Centroid(2));
            cropped_x = uint16(s(hand_cluster_num).BoundingBox(1))+1;
            cropped_y = uint16(s(hand_cluster_num).BoundingBox(2))+1;
            cropped_w = uint16(s(hand_cluster_num).BoundingBox(4))-2;
            cropped_h = uint16(s(hand_cluster_num).BoundingBox(5))-2;
            % apply gaussian blur to remove noise
            PSF = fspecial('gaussian',7,10);
            L = imfilter(L,PSF,'symmetric','conv');
            % cropped region of interest from main image
            M = L( cropped_y : (cropped_y+cropped_h), cropped_x: (cropped_x+cropped_w) );
            % trim leading and leading zeroes
            M(:, logical(sum(abs(M)) == 0)) = [];
            cropped_h = size(M,1);
            cropped_w = size(M,2);
            non_zeroes = find(sum(abs(L))~=0);
            cropped_x = non_zeroes(1);
            % fill holes to remove flares
            M = imfill(M, 'holes');
           % get a shrunk sample from cropped image
            N = imresize(M, 0.25);
            [sample_height sample_width] = size(N);
            % clip excess pixels
                N = N(1:50,:);
               N = N(:,1:50);
            [sample_height sample_width] = size(N);
            %pad both sides
            htopad = double(25 - ceil(sample_height/2));
            wtopad = double(25 - ceil(sample_width/2));
                N = padarray(N,[htopad wtopad]);
                %for uneven padding, add padding to end
                [sample_height sample_width] = size(N);
                htopad = 50 - sample_height;
                wtopad = 50 - sample_width;
                N = padarray(N,[htopad wtopad],0,'post');
                disp('Padding not performed...');
     %this two lines is for debugging purpose only
            hold on;
                plot(centroid_x, centroid_y, 'bo');
               rectangle('Position', [ cropped_x cropped_y cropped_w cropped_h ], 'EdgeColor','y');
            hold off;
            croppedImage = N;

Tags for this Thread

What are tags?

A tag is like a keyword or category label associated with each thread. Tags make it easier for you to find threads of interest.

Anyone can tag a thread. Tags are public and visible to everyone.

Contact us